Abstract
The widespread use of chlorine pre-oxidation in water purification has been limited in several countries owing to the production of carcinogenic byproducts when combined with naturally occurring organic matter. This study investigates the efficient use of potassium permanganate (KMnO4) pretreatment and coagulation enhancement as particle size and molecular weight distribution controlling parameters. KMnO4 pretreatment significantly reduced the apparent molecular weight of humic acid due to KMnO4 reduction and the continuous generation of manganese dioxide (MnO2) formed in situ under neutral and alkaline conditions. The MnO2 formed in situ had adsorption characteristics that enabled it to form large and stable flocs with the hydrolysis products of aluminum sulfate. However, under acidic conditions, KMnO4 pretreatment exhibited strong oxidation characteristics due to Mn(VII) reduction to Mn(II), and the mean particle floc size was the same as without KMnO4 pretreatment. Overall, KMnO4 pretreatment is a useful alternative strategy for traditional pre-oxidation using chlorine and a good coagulant enhancement agent in neutral and basic media.
HIGHLIGHTS
pH is crucial in the KMnO4 pretreatment mechanism.
Under acidic conditions, KMnO4 pretreatment results in the same mean particle floc size as alum alone.
The in situ formed MnO2 is incorporated with the alum hydrolysis product, forming larger flocs.
KMnO4 pretreatment under optimal conditions results in a significant reduction in the molecular weight distribution of humic acid.
Graphical Abstract
INTRODUCTION
Egypt depends on conventional treatment for surface water purification from the Nile River and its branches. The treatment is a series of processes that include coagulation, flocculation, sedimentation, and filtration and is typically followed by full-scale disinfection (Zouboulis et al. 2008). Water treatment plants (WTPs) are often preceded by a pre-oxidation step. Chlorine has been extensively used in developing countries for WTPs to inactivate harmful pathogens and prevent algae accumulation in pipes, tanks, and filters. However, the pre-chlorination step has some limitations because chlorine can react with natural organic matter (NOM) to form various harmful carcinogenic disinfection byproducts (DBPs), such as trihalomethanes (THMs) and haloacetic acids (HAAs) (Li & Mitch 2018). As a result, several water regulatory agencies have lowered the maximum allowable levels of THMs and HAAs in drinking water (USEPA 2003).
One possible strategy to reduce DBP levels in treated water is for WTP operators to perform the chlorination step after coagulation, sedimentation, and filtration, which significantly reduces THM formation, sometimes by more than 50% (García & Moreno 2009). However, eliminating the pre-chlorination step may not be a viable option, as it has been used for controlling biological growth, taste, and odor during water treatment processes (Lapsongpon et al. 2017).
The use of alternative pre-oxidants, such as ozone, chlorine dioxide, UV, and potassium permanganate (KMnO4), instead of chlorine has been investigated in many studies (Xie et al. 2016). KMnO4 has long been used for removing iron and manganese by oxidation and precipitation. Recently, KMnO4 has received much attention for its use in treating NOM and removing algae. KMnO4 utilization has the advantage of lower installation costs, easy application, and production of nonhazardous byproducts (Yu et al. 2011; Tian et al. 2013; Phatai et al. 2014).
Many researchers have discussed the efficiency of KMnO4 in removing DBP precursors. While some authors attributed its enhancement to the chemical oxidation reaction, others believed it was physical adsorption on the in situ formed manganese dioxide (MnO2).
Hidayah & Yeh (2018) reported that KMnO4 pre-oxidation caused the breakdown of high-molecular-weight organics into low-molecular-weight organics. They also revealed that increasing KMnO4 dosage could decrease aromatic matter, as evidenced by 23 and 28% reductions in UVA254 and SUVA254 values, respectively. Moreover, DBP formation potential in terms of THM formation potential and HAA formation potential decreased by approximately 15 and 23%, respectively.
Another approach was studied by Ma et al. (2001), who performed pre-oxidation with KMnO4 to improve surface water coagulation using high organic content. They found that the MnO2 formed in situ during KMnO4 reduction significantly improved the coagulation and sedimentation processes. Manganese dioxide can adsorb naturally occurring organic materials through surface binding to form larger particles, increase the floc density, and improve the particle settling velocity (Liu et al. 2012; Lin et al. 2013; Cui et al. 2020).
Previous studies have discussed NOM removal by KMnO4 pre-oxidation at specific pH values (Zhang et al. 2009; Liu et al. 2012; Xie et al. 2016) without analyzing particle size and precursors' molecular weight distribution and their relationship to KMnO4 efficiency in coagulation enhancement. Therefore, the current study tries to fill the knowledge gap and complete the picture of using KMnO4 as an alternative pre-oxidant to enhance the removal of DBP precursors by determining the reaction pathway at different pH values using floc particle size and molecular weight distribution. Different modes of action are identified (oxidation or adsorption), and detailed analyses for controlling parameters are presented.
MATERIALS AND METHODS
Chemicals
Humic acid (HA): A stock solution was prepared by dissolving 1 g HA (Fluka) in 1,000 ml of 0.1 M NaOH, which was then filtered through a 0.45 μm filter membrane and stored at −4 °C in the dark. The stock solution was diluted, and the total organic carbon (TOC) was measured and recorded.
The stock solutions used in this study included KMnO4 solution (0.5 g/L), Al2(SO4)3·18H2O coagulant (0.5 g/L), and kaolin (50 g/L). All chemicals were of reagent grade, and all solutions were made using deionized water.
Synthetic water preparation
The synthetic water samples comprised HA, kaolin, and tap water from Shoubra El-Kheima WTP and were used to simulate actual surface water contaminants. Table 1 lists the properties of the synthetic water.
Synthetic water properties
No. . | Parameter . | Value . |
---|---|---|
1 | TOC content | 5 mg/L |
2 | UV254 absorbance | 0.700 ± 0.002 |
3 | pH | 8.25 ± 0.10 |
4 | Turbidity | 30.0 ± 0.5 NTU |
No. . | Parameter . | Value . |
---|---|---|
1 | TOC content | 5 mg/L |
2 | UV254 absorbance | 0.700 ± 0.002 |
3 | pH | 8.25 ± 0.10 |
4 | Turbidity | 30.0 ± 0.5 NTU |
Preliminary determination of alum and KMnO4 dosage
The alum working dose was predetermined using a standard jar testing apparatus (Hach) with a six-paddle stirrer in 2-L square plexiglass beakers. It was determined based on the turbidity removal efficiency of the artificial samples. The working KMnO4 dose was estimated based on the turbidity removal efficiency and the permissible residual Mn(II) according to Egyptian drinking water quality regulations.
Experimental setup
The jar test experiments were conducted on the synthetic samples at room temperature of 25 °C ± 1 °C. The predetermined KMnO4 dose was mixed at 225 rpm for 5 min, followed by the addition of the determined alum dose at the same rate for 2 min, and then the mixing speed was lowered to 35 rpm for 15 min. The solution was allowed to settle for 15 min. All samples were analyzed for turbidity using a Hach 2100 N turbidimeter (USA).
Size exclusion chromatography
The variation in HA concentration and molecular weight distributions were analyzed by gel permeation chromatography (PL-GPC 50 with integrated RI, Agilent Technologies, USA) during the jar test experiment after KMnO4 pretreatment, flocculation, and sedimentation. The samples for GPC analyses were filtered through glass fiber filter paper with a pore size of 0.45 μm before analysis.
Floc size distributions
The development of floc size distributions during the jar test experiments were continuously monitored using a laser particle size analyzer (Mastersizer 2000, Malvern Instruments, UK) by drawing water through the optical unit of the instrument and back into the jar test beaker again with a peristaltic pump on the return tube. The inflow and outflow tubes were positioned in the jar opposite one another at a depth just below the surface by 4 cm. Mean floc size measurements were taken every 30 s during the jar test experiments and logged onto a computer (Xu et al. 2016).
RESULTS AND DISCUSSION
Determination of the optimal doses of alum and KMnO4
KMnO4 has oxidation characteristics that might degrade HA into low-molecular-weight fragments, which may be removed by subsequent coagulation. In addition, the MnO2 formed in situ has adsorption characteristics that may be the factor responsible for HA removal, or perhaps both mechanisms are working together (Zhang et al. 2009; Liu et al. 2012).
Dynamic analysis of floc size during alum coagulation jar test with KMnO4 pretreatment
Time course of mean floc size development with and without permanganate pretreatment under different pH values.
Time course of mean floc size development with and without permanganate pretreatment under different pH values.
Under acidic conditions, the mean floc diameter increased rapidly at 2 min, reaching a maximum with a mean floc diameter of approximately 100 μm and remained constant after 5 min flocculation both with and without KMnO4 pretreatment. This effect occurred because of KMnO4's strong oxidation properties and Mn(VII) reduction to Mn(II) (Kao et al. 2008; Li et al. 2012), which has no effect on the mean floc diameter size.
The largest mean floc diameter was formed at pH 7.0 with KMnO4 pretreatment, which started at 4 min of flocculation and gradually increased because of the continuous generation of colloidal MnO2, which combined with the hydrolysis products of alum and aggregated into larger flocs. After 15 min of flocculation, the maximum steady value was reached at approximately 105 μm, compared to 63 μm in the case without KMnO4 pretreatment.
The coagulation jar test with the selected alum dose at pH 8.5 did not produce flocs. However, KMnO4 pretreatment induced floc formation after 8 min, and the mean floc particle size increased continuously owing to the continuous formation of colloidal MnO2 during KMnO4 reduction, reaching 100 μm after 18 min.
Characterization by GPC
Molecular weight (MW) distributions of (a) untreated original artificial HA samples at different pH values and treated HA using the KMnO4/alum coagulation process, (b) pH 5.5, (c) pH 7.0, and (d) pH 8.5.
Molecular weight (MW) distributions of (a) untreated original artificial HA samples at different pH values and treated HA using the KMnO4/alum coagulation process, (b) pH 5.5, (c) pH 7.0, and (d) pH 8.5.
The artificial HA comprised two main bands with different molecular size fractions at all selected pH values. As shown in Figure 4, the percentages of the two peaks decrease with lowering the pH of the original HA because HA with a few negative charges was removed more by self-aggregation at lower pH values (Hakim et al. 2019).
Before the addition of alum, the chromatogram of HA after 5 min of KMnO4 pre-oxidation showed a decrease in the organic matter content under neutral and basic conditions (Figure 4(c) and 4(d)). This is because the MnO2 formed in situ has adsorption ability and may adsorb HA fractions (Xie et al. 2013). In contrast, the organic contents increase under acidic conditions because HA organic compounds are degraded to low-molecular-weight compounds during KMnO4 reduction to Mn(II), proving that the oxidation mechanism prevails (Figure 4(b)).
Furthermore, under the same alum dosage of 1.0 mg/L as Al(III), the apparent molecular weight was reduced, and the high-molecular-weight compounds were more easily removed by the subsequent alum coagulation process than the low-molecular-weight ones. Notably, coagulation with KMnO4 pre-oxidation caused a greater reduction in peak height and area for all types of molecular weight organic compounds than coagulation without KMnO4 pretreatment, especially under basic and neutral conditions. This effect occurred because the MnO2 formed in situ can adsorb HA and then interact with the hydrolysis product of alum to form large flocs with higher specific gravity and higher settling velocity, resulting in improved HA removal (Ma et al. 2001).
CONCLUSION
This study investigates KMnO4 as a feasible alternative pre-oxidant and coagulant enhancement agent. pH is a critical factor affecting KMnO4 behavior in the entire mechanism and process, which could be either only strong oxidation under acidic conditions or mild oxidation and adsorption through the in situ formed MnO2 at neutral and basic media in proportion. The molecular weight distribution was significantly reduced using KMnO4 under neutral and basic conditions, indicating that HA was broken down into lower-molecular-weight fragments that could be easily adsorbed by coagulant hydrolysis products and in situ formed MnO2. Furthermore, particle size analysis revealed that the continuous formation of colloidal MnO2 under neutral and basic conditions allows it to combine with alum hydrolysis products and aggregate into larger flocs, enhancing the coagulation process.
Overall, KMnO4 is an excellent strategy and alternative to pre-oxidation, particularly in areas prone to organic loads, to overcome the issue of DBP formation. After evaluating its performance on natural surface water samples, we recommend using KMnO4 pre-oxidants on WTPs with a high organic load.
ACKNOWLEDGEMENT
The authors proudly acknowledge Dr. Ahmed Gamal Yahia for his contribution in the graphical abstract.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.